PBE.fit.f: Fit a PB vs E relationship

In duplisea/ccca: Climate Change Conditioned Advice for Sustainable Fisheries Exploitation

Description

The various model fits: polynomial "poly", GAM "gam", adaptive GAM "gam.adaptive", various scam fits and resamples from the PB values "avg" can be chosen. avg just fits a linear model with slope = 0 and then resamples the residuals which is effectively the same as just sampling the P/B values directly, i.e. it does not force a relationship between P/B and E and therefore the future is just a resampling of the past. scam (shape constrained additive models) fits force certain characteristics in the shape such as monotonicity, convex, concave, increasing or decreasing.

Usage

PBE.fit.f(PB, model.type, knots, poly.degree)

Arguments

PB	the data and model fit coming from applying the PB model (PB.f)
model.type	the kind of model to fit ("poly", "gam", "gam.adaptive","avg","mpi","mpd","cx","cv","micx","micv","mdcx","mdcv"
knots	the number of knots for adaptive GAM
poly.degree	the degree of the polynomial to fit

See Also

[mgcv::gam()], [mgcv::smooth.terms], [mgcv::scam()], [mgcv::shape.constrained.smooth.terms], [lm()]

duplisea/ccca documentation built on April 17, 2021, 2:31 a.m.